454 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1921. 



light, plenty of time, and due care, in determining the shade. In 

 general this method is preferable to that of comparing the hair with 

 given standards, for that takes longer and among such a mixed popu- 

 lation as ours we would never have enough standards. It is true that 

 it is not easy in such a visual method to get rid of all personal equa- 

 tion, but the amount of such an equation may be very much reduced 

 and be rendered practically insignificant by due instruction or under- 

 standing of the subject, with practice. The final classification of the 

 shades is not arbitrary. We begin with the safe units of " black, " 

 decidedly " light, " and unmistakably red. This leaves a large cate- 

 gory of intermediate grades, all of which fall, however, into three 

 subdivisions, namely, light brown (not blond) , medium (or " medium 

 brown"), and dark (or "dark brown"). A large majority of cases 

 will readily and unmistakably be placed in one or another of these 

 classes by every properly instructed observer. This will leave, as pos- 

 sible sources of error, only the transitional shades, for there are be- 

 tween none of the colors any definite lines of demarcation. These 

 cases, with a careful student, will amount to approximately 10 per 

 cent with the blonds, 20 per cent with the light browns (not blond) , 

 mediums, darks, and reds, and 5 per cent with the blacks. When we 

 add to this that by the law of chance, other things (such as the train- 

 ing of the observers, etc.) being equal, as many of the " uncertains " in 

 each category will be recorded right as wrong, and that those re- 

 corded wrong in one class will be counterbalanced by the wrongs of 

 the next, it may be seen that unless there is a lack of due instruction, 

 negligence, or the development of some special bias on the part of 

 an observer, his records on any large series of individuals will be 

 substantially correct and comparable with those of all other similarly 

 instructed and careful workers. That this is so may be shown in 

 our series in Virginia. In a camp of the U. S. Corps of Engineers, 

 near Charlottesville, after due initiation the work was left in the 

 hands of Dr. Robert Bennett Bean, of the University of Virginia. 

 The results, except for a slight difference which developed in record- 

 ing the eye colors, were practically identical with those of the author 

 as far as the latter applied to the same territory. 



The study of hair color among the Old Americans fully confirms 

 previous observations on the change in the color of the hair with age. 

 Except in those with the darkest shades the hair in general shows 

 from infancy on to adult life and in many cases even through a part 

 of the adult life, a progressive darkening. The lightest hair in an 

 infant may thus eventually become light, medium, or even fairly 

 dark-brown — though not black. Even the red hair darkens or loses 

 its purity. The golden also is unstable. A small series of near-adults 

 found by the writer among the teachers shows, as will be shown later 



